Potu Gal

Canadian Joumal of Administrative SciencesRevue canadienne des sciences de l'administration24: 2-14 (2007)Published online in Wiley Interscience (www,interscience,wiley,com), DOI: 10,1002/CJAS,l

®WILEY

iTiterScience*

Conformity and Diversity of Accountingand Financiai Reporting Practices inPortuguese Locai GovernmentJoao Baptista da Costa CarvaihoPedro J. Camoes*The University of Minho

Susana Margarida JorgeThe University of Coimbra

Maria Jose FernandesThe Polytechnical Institute of Cavado and Ave

AbstractPortuguese governmental accounting has evolved fromessentially a cash-based budgetary accounting system toan accrual-based financial and cost accounting system.Evidence from financial reports shows a great level ofdiversity amongst municipalities' voluntary compliancewith the new accounting rules. Using a sample of Portu-guese municipalities in the year 2003 we calculate andanalyse the levels of compliance with the practicesrequired by law and document the diversity in compli-ance across municipalities. The differences acrossmunicipalities are explained by some fundamentalfactors: size, financial conditions, urban characteristics,and diffusion across neighbouring municipalities. Weshow an unexpected effect of size. Larger municipalitiescomply less with accounting standards. Organizationalcomplexity, conservative practices, and aversion tochange may explain this result. Copyright © 2007 AS AC.Published by John Wiley & Sons, Ltd.

JEL classification: M41

Keywords: Portugal, local government accounting,financial and budgetary reporting, compliance, diversity

ResumeLa comptabilite du gouvernement local au Portugal aconnu des mutations. On est ainsi passe de la compt-abilite de tresorerie au systeme de comptabilite analyt-ique et de comptabilite d'exercice. Toutefois, d'apres Iesrapports financiers, Ies municipalites ne se conformentpas toujours de la meme fagon aux nouvelles regiescomptables. A partir d'un echantillon de municipalitesportugaises constitue en 2003, nous calculons et analy-sons Ies niveaux de conformite aux pratiques exigees parla loi et la diversite entre ces pratiques. Les differencesentre les municipalites s'expliquent par des facteursessentiels tels que la taille, les conditions financieres, lescaracteristiques urbaines et la diffusion entre munici-palites voisines. Notre etude met en evidence I'infiuenceinsoupgonnee du facteur taille. Les municipalites degrande taille semblent avoir plus de difficulte a appliquerles regies comptables. La complexite organisationnelle,les pratiques conservatrices et I 'aversion pour le change-ment sont quelques-uns des facteurs responsables decette situation. Copyright © 2007 ASAC. Published byJohn Wiley & Sons, Ltd.

Mots-cles : Portugal; comptabilite du gouvernementlocal; rapport financier et budgetaire; conformite;diversite

Sponsored by POCTI/CPO/58391/2004,*Please address correspondence to: Pedro J, Camoes, Economics and Management School, The University of Minho, Campus de Gualtar, 4700 Braga,Portugal, Email: pedroc@eeg,uminho,pt

Copyright © 2007 ASAC, Published by John Wiley & Sons, Ltd,Can J Adm Sci

24(11, 2-14

CONFORMITY AND DIVERSITY OF ACCOUNTING AND FINANCIAL REPORTING PRACTICES DA COSTA CARVALHO ET AL

Worldwide public administration has undergoneseveral changes in the last twenty-five years that havebrought public sector practices closer to those of privatebusiness. These changes have come to be known as the'New Public Management' movement. Specifically withregard to governmental accounting and financial report-ing, accrual based financial systems and performancemeasurement systems have been introduced alongsideGeneral Accepted Accounting Principles (GAAP).

Portuguese governmental accounting has undergoneconsiderable developments since the early 1990s. Ithas evolved from an essentially cash-based budgetaryaccounting system to an accrual-based financial and costaccounting system. A single double-entry system thatintegrates budgetary, financial and management subsys-tems has been in use since 1997. This system is supportedby an Official Plan for Governmental Accounting(OPGA)' and is very similar to the plan currently usedin business accounting. This plan has been adapted to themain subsectors of public administration, namely educa-tion (Instruction 794/2000), health (Instruction 898/2000),social security (Law-decree 12/2002), and local govern-ment (Official Accounting Plan for Local Government[OAP-LG], Law-decree 54A/99).

However, in practice, there is great diversity in theway in which the plan has been applied across munici-palities. This calls into question governmental account-ing reform aimed at ensuring standardization anduniformity. In fact, most local governments do not fulfilthe main requirements of the new system. Moreover, thediversity in the stages of implementation tends to behigher for financial accounting practices than for budget-ary accounting practices.

Despite the obligation since 2002 that all munici-palities present their accounts in accordance with theOAP-LG, no audits have occurred. Audit is not requiredin Portugal, so compliance is essentially voluntary.Because there is no enforcement mechanism and nopenalty for noncompliance, municipalities choose whichlegal requirements to apply and which to ignore.

The main objective of this paper is to evaluate atwhat stage Portuguese municipalities are with respectto implementing legal requirements of the OAP-LG. Inparticular, we assess the extent to which municipalitiesare complying with the new standards and documentcharacteristics that explain differences in level of com-pliance across municipalities. Although several recentstudies on the subject exist (Carvalho & Jorge, 2003;Femandes & Carvalho, 2004; Jorge, 2003; Jorge, 2004),this body of research has not assessed the determinantsof compliance.

Our paper is organized as follows: The next sectiondetails the context of Portuguese local government and

the main aspects of the accounting reform. Followingthat, we calculate the degree of uniformity in accountingpractices across municipalities throughout Portugaland the extent to which these municipalities are com-plying with the requirements of the new system. Wethen analyse determinants of the levels of compliance.Finally, we offer explanations for why some municipali-ties are in compliance with OAP-LG while others arenot, and conclude by offering suggestions for futureresearch.

Portuguese Local Government Context and Accounting

Portuguese local governments are comprised of par-ishes, municipalities, and other forms of local territorialorganizations that may be created by law, such as metro-politan areas, district assemblies, and associations ofparishes and municipalities. Of these, municipalities arethe most important from an administrative and financialperspective.

There are currently 308 municipalities (continentalarea and autonomous regions of Madeira and Azores)representing approximately 9% of the total publicexpenditure and 5% of the Gross Domestic Product.Local authorities and municipalities in particular aimto satisfy the interests of a certain geographic territory.Their main responsibilities cover the followinggovernmental domains: environment; culture; leisure;sports; economic development; civil protection; urbantransports; territorial arrangement; social housing andsupport; and heritage preservation. Their political orga-nization is comprised of an executive committee and alocal (deliberative) council, which follows the centralgovernment model. Their revenues essentially comefrom:

- grants from the Central Government and EuropeanUnion (earmarked and non-earmarked transfers);

- municipal (property and income) taxes;- loans;- fees, rates, prices, and fines;- property revenues (rents and assets sales, heritages

and donations, income from financial participation inmunicipal business companies, and other financialassets).

Municipalities and other local government entitiesare financially autonomous, meaning that they managetheir own budget, finances, property, loans, and treasury.The services they provide have budgetary independencefrom the State's budget. Central government interventionis merely inspective of legal compliance, particularly

Copyright © 2007 ASAC. Published by John Wiley & Sons, Ltd.Can J Adm Sci

24(n. 2-14


with respect to: multi-annual plans for investment; budgetand budgetary execution; accounting; revenue genera-tion and collection; authorization and payment of ex-penditures; debt; property management; and fiscalobligations.

With respect to budgeting and accounting, the imple-mentation of a true local government accounting systemoccurred only after the restoration of the democraticregime^ and the 1976 Constitution's recognition of thepolitical merit of local government. After this historicalturn, local government's political and financial autonomysuggested the need for a specific accounting regime andin 1979, the first Local Finances Law was approved. Thecurrent Local Finances Law caused considerable changesto the previous accounting system. Of significance wasthe introduction of accrual-based, double-entry financialand cost accounting systems, which replaced cash-based,single-entry budgeting. The newer system is similar tobusiness accounting and includes cost accounting for theso-called "municipalized services" - autonomous andbusiness oriented units within municipalities - and fed-erations of municipalities. The Local Finance Lawrequires the local government accounting regime to applyuniform and standardized systems. The most importantfeatures of the new local government accounting frame-work are:

- a single system that records and integrates operationsin three subsystems - budgetary, financial, and eco-nomic - beginning with approval and execution ofbudgets and extending through to the allocation ofcosts;

- accounting principles that align closely to businessGAAP and follow specific budgetary rules;

- the valuation of fixed assets at historical cost,allowing for valuation at net realizable value orreplacement cost only in exceptional circumstances;

- calculation of costs of services provided, allowingcosts by functions, activities, and departments forinternal management purposes;

- complementary legislation for fixed assets actualiza-tion (Instruction 671/2000) and an economic budget-ary classification for revenues and expenditures(Law-decree 26/2002);

- financial statements such as balance sheets (disclosingtotal assets recorded at book value, stocks, debts, andother liabilities) and operating statements (disclosinginformation on costs and revenues by category andannual results of operations) that are reported togetherwith budgetary execution statements;

- public domain assets (infrastructure, cultural, and heri-tage assets) to be listed in the balance sheet and com-pulsory depreciation if appropriate;

- assets acquired under a financial leasing system to beimmediately recognized as fixed assets;

- information to account for each entity's budgetary,financial, and economic situation, as well as to supportdecision-making (management);

- a legal framework of standards produced bynational oversight bodies, namely the Ministry ofFinance and The Court of Accounts (SupremeAudit Office),̂ supported by the Public Administra-tion Accounting Standardization Commission(Law-decree 68/98).

Moreover, the three mandatory accounting subsys-tems that currently exist in local administration havedifferent features and objectives: budgetary accountingaims at controlling the budget execution, financialaccounting reports the entities' economic and financialsituation, and cost accounting calculates the cost of ser-vices, activities, or departments (Fernandes & Carvalho,2004).

The new system presents some drawbacks (Jorge,2003; Fernandes & Carvalho, 2004), with the most sig-nificant including:

- The "patrimonial perspective" (listing of all property)has created difficulties in tracking inventory andvaluing public domain assets. In particular, historicalcost accounting fails to capture any artistic or culturalvalue attaching to these assets.

- Full costing, required to be used in cost accounting,has raised questions related to cost allocation, particu-larly the allocation of indirect costs.

- The absence of consolidation rules: considering thatmunicipalities have increased their peripheral sector,an overall picture of the economic-financial positionof the entity as a whole (namely embracing municipal-ized services and municipal business enterprises) isincreasingly imperative.

- There is a lack of further development in manage-ment accounting and performance reporting. Im-proving municipal management and perfor-mance measurement would contribute to assessingeconomy, efficiency, and effectiveness in publicservices.

- Budgets and budgetary accounting are still preparedon a cash (or modified cash) basis. Because the budgetis the most important political instrument in local gov-ernment, control of its execution is central to dailymanagement. The continued use of cash reporting maybe because politicians potentially have problemsunderstanding accrual basis accounting and thereforedo not see the utility of the balance sheet and resultsstatement.


24(11. 2-14


Empirical Assessment of Compliance and Diversitythroughout Portugal

Data

The data we used are based on a random samplerepresenting 57% of Portuguese municipalities - 175 outof 308. We collected the data from the 2003 municipalannual accounts from the archives of the Supreme AuditOffice of the Court of Accounts and from the municipali-ties themselves where necessary. Since residentpopulation is one of the factors used to determine inter-governmental grants (Local Finance Law 42/98), wegroup municipalities according to population size asfollows: small < 20,000 residents; medium 20,000 to100,000 residents; large > 100,000 residents.

Analyses

Compliance. We begin by computing a total com-pliance index that measures the level of implementationof the new accounting system while taking into accountmunicipality size. It allows us to compare practicesreflected in the annual accounts, namely, financial andbudgetary statements and notes (with standards such asthose established in the OAP-LG as well as in The Courtof Accounts Resolution 4/2001).

The analysis is limited to a given set of issues thatcombines information on budgetary items with the majorinnovations brought in by the new accounting framework(Allen & Sanders, 1994; Christiaens, 1999; Pina Mar-tinez & Torres, 1995; Torres & Pina Martinez, 2003).Twenty different issues were selected and groupedaccording to practices of the three accounting subsys-tems: budgetary, financial, and management accounting(see Table 1). Given that 6 of these 20 items are common

Figure 1.Municipalities according

Number ofMunicipaiities

180160140120-100806040-20-0-1

I/a 1Ml 4

Small

to size

^ .̂103

IkMedium

Size

11

• Population• Sample

Large

to more than one subsystem, there are a total of 26 itemscompared across systems.

Municipal practices and information are scoredagainst these items, scoring 1 when they are conductedand 0 when they are not. We assumed that all items areequally important, therefore, all have the same weight inthe index: "*

Index = ^ Pi,

where p, represents the item in question (m < 26). Eachmunicipality may reach a maximum compliance index of26 points. From this we calculate the global complianceindex for the whole sample. This is the mean of allmunicipal compliance indexes. We also report a globalindex of partial compliance documenting which practicesand information subsystems have a greater level of com-pliance. In order to assess possible differences in compli-ance among groups, both indexes are calculated takinginto consideration municipality size.'

Table 2 shows the partial compliance indexes foreach of the three accounting subsystems grouped accord-ing to size.

Compliance is higher with practices and informationwithin the budgetary accounting subsystem. In fact, theaverage level of compliance with budgetary accountingparameters is 87% for our sample as a whole. In otherwords, all entities comply with 6 out of 7 of the items onaverage. The indexes for financial and managementaccounting systems are only 54.5% and 26.7% respec-tively. With regard to the subsystems, compliance isgreater in the financial accounting subsystem in medium-sized municipalities, where we observe a compliancelevel of 58%. However, this partial compliance indexdoes not vary much when we turn to small or large enti-ties. It is very similar across the subsystems regardlessof municipalities' size.

Compliance with management accounting parame-ters is slightly higher for large and medium municipali-ties, both of which comply with 29% of the parameters,as compared to their smaller counterparts, who complywith 25% of the parameters. The similarity of theseresults casts doubt as to whether size truly has aneffect on whether municipalities comply with practicesand information in the management accountingsubsystem.

The %^ test results, shown in Table 3 suggest inde-pendence between size of municipality and compliancewith respect to budgetary and management accountingsubsystems. This is not the case with the financialaccounting system where the x^ test results demonstratea significant association between size of municipality andcompliance (p < .05).


24(0. 2-14

CONFORMITY AND DIVERSITY OF ACCOUNTING AND FINANCIAL REPORTING PRACTICES DA COSTA CARVALHO ET AL.

Table 1Information and Practices of the New Local Government Accounting System

Subsystems Parameters No. of Municipalitiescomplying

Budgetary accounting

Financial/ patrimonialaccounting

Management accounting

1. Expenditure execution 1752. Revenue execution 1753. Economic classification 1754. Functional classification 1115. Degree of budgetary execution (budget/actual) 1516. Liabilities (short and long term) 1697. Long-term investment projects 1118. Public domain assets 1649. Operational assets 169

10. Public domain assets depreciation 8411. Operational assets depreciation 15012. Deferred revenues 13713. Accrued expenses 9014. Provisions (assets) 2515. Provisions for contingencies 316. Receivables 15917. Liabilities (short and long term) 16918. Costs of goods/services, functions and departments 019. Results of goods/services, functions and departments 020. Long-term investment projects 11121. Accounting policies (valuation criteria, derogated establishments, etc.) 7522. Public domain assets depreciation 8423. Operational assets depreciation 15024. Costs of goods/services, functions and departments 025. Performance measurement (3E) 026. Results of goods/services, functions and departments 0

Diversity. In this section, we calculate a concentra-tion index that measures diversity in municipal practices.We calculate this index for the whole sample and for thetwo systems analyzed above with the greatest compli-ance index (budgetary and financial) according to munic-ipality size. We group these according to three intervals:below 33.3%, between 33.3-66.6%, and above 66.6%.These intervals represent three alternatives of compli-ance against which the relative frequencies are weighted;a greater relative frequency implies a greater weight. Therelative frequency represents the percentage of munici-palities that comply with a percentage of items withinthose intervals.

We rely on the Herfindahl-H index:

where Pi is the relative frequency of the interval and n isthe number of intervals. Values vary between 0 and 1.The lower the concentration index, the greater

the diversity among municipalities' practices andinformation.

Table 4 shows that the concentration indexes arehigher for budgetary than for financial accounting prac-tices, implying that there is less diversity in compliancewith budgetary (as opposed to financial ) parametersacross municipalities of differing sizes. The sample Hvalue is 0.86, but it is even higher for medium-sizedmunicipalities (around 0.91). This seems to indicate thatconcentration might somehow be related to size. However,regardless of size, more than 90% of the entities analyzedconcentrate their practices on more than 66.6% of theparameters.

With respect to financial accounting, the concentra-tion index is in the vicinity of 0.6 for the whole sampleand is slightly higher for large municipalities (0.63),which indicates that concentration may be affected bysize. Compared to the results for the budgetary account-ing subsystem, the concentration index is lower regard-less of municipality size. Approximately 75% of entities


24(11. 2-14


Table 2Partial Compliance

Number of total parameters per subsystem Number of municipalities

Sample

0001

12511695

6.0971(87.10%)

10323

142333342727

8000

7.6343(54.53%)

256684000

1.3371(26.74%)

15.0685757.96%16

Large

0000262

12

6.0909(87.01%)

000102245431000

7.7272(55.19%)

01210000

1.4545(29.09%)

15.2727358.74%16

Medium

00003

197

33

6.129(87.56%)

0010027

121510105000

8.1129(57.95%)

13229000

1.4516(29.03%)

15.6935560.36%16

Small

00017

267

50

6.0769(86.81%)

10213

1014171413142000

7.2857(52.04%)

242245

000

1.2307(24.62%)

14.5934156.13%14

Budgetary accounting01234567

Weighted mean

Financial accounting01234567891011121314

Weighted mean%

Management accounting0I2345

Weighted mean

Total compliance indexWeighted mean%Mode

Notes: 1. Values in columns 2-5 tell us how many parameters (budgetary, financial, or management) are partially complied with byPortuguese municipalities (e.g., with respect to budgetary parameters, 95 municipalities comply with all 7 budgetary parameters - row8, column 2); 2. There are maximums of 7 budgetary parameters, 14 financial parameters, and 5 management accounting parameters;3. As an illustration of the computation of the compliance percentages, the 87.1% compliance with budgetary accounting parametersover the entire sample is computed as follows: (0*1 + 1*0 + 2 * 0 - H 3 * 1 -1-4*12 + 5*51 +6*16 + 7*95)/l,225. The numerator ofthis calculation indicates that no municipalities complied with 2 or fewer parameters, 1 municipality complied with 3 parameters, 12municipalities complied with 4 parameters, 51 municipalities complied with 5 parameters, 16 municipalities complied with 6 parameters,and 95 municipalities complied with all 7 parameters. In order to compute the % compliance, we divide by the perfect compliance scoreof 1,225 (175*7), which would be attained if all 175 municipalities complied with all 7 budgetary accounting parameters.


24(0. 2-14


concentrate their practices in the median interval, com-plying with a number of items somewhere around 50%.

The x^ test results presented in Table 5 show that wecannot reject the null hypothesis in the case of budgetaryaccounting, indicating that size does not seem to affectthe diversity of accounting parameters analyzed. Thesame is not so in the case of financial accounting wherewe can reject the null at a significance level of 5%.

Determinants of Compliance

Our analyses of compliance and diversity raise thefollowing questions: Why do some local governmentsimplement the new accounting system while others donot? Why do some municipalities have high levels ofcompliance and others low levels? What accounts for thisdifference in the level of compliance?

Although bigger cities have more facilities to adapttheir accounting system to the reforms, size of munici-

Table 3Chi Square Tests Results for Partial Compliance

SubsystemsBudgetary accountingFinancial accountingManagement accounting

0.12867.33152.1901

Note: (*) X = 5.99 with 5% significance level and 2dfs.

pality does not seem to consistently explain the differ-ences in implementation of generally accepted accountingprinciples in local administration. Further empirical tests,particularly a more robust test of the infiuence of sizeand of other possible explanations, are necessary. Theliterature on the determinants of disclosure both in theprivate sector and in the public sector is abundant (Allen& Sanders 1994, Christiaens 1999, Street & Gray 2002).Agency theory, political costs theory, signalling theory,legitimacy and institutional theory, property costs theory,and contingency theory are just a few of the theoreticalframeworks used to study disclosure.

When analysing the effect of several factors on aninformation disclosure index in US municipalities, Allenand Sanders (1994) concluded that city size (measuredusing population) does not affect disclosure, although thefindings seem to indicate that an association might existwhen size alone is considered in the absence of otherexplanatory variables. In turn, Christiaens (1999) carriedout a regression analysis considering, among other deter-minants, the effect of size on a compliance index ofaccounting practices for Flemish municipalities. Hisfindings support a significant and positive effect on thecompliance index confirming the hypothesis that munici-pal size is a significant determinant of the level of com-pliance. Following these studies, we test the hypothesisof the influence of size including population. Followingthe standard statistical recommendation, we applied alogarithmic transformation to measure population size(log population).

Table 4Concentration Index: Diversity of Practices and Information

Subsystems I of parameters complied with Relative frequencies

Sample Large Medium Small

Budgetary accountmg

Financial accounting

<33.3%>33.3%<66.6%>66.6%H<33.3%>33.3%<66.6%>66.6%H

0.0000.074

0.9260.86250.0510.749

0.2000.603

0.0000.091

0.9090.83470.0450.773

0.1820.6322

0.0000.048

0.9520.90790.0160.742

0.2420.6093

0.0000.088

0.9120.83960.0770.747

0.1760.5952

Notes: 1. This table documents the percentage of parameters with which municipalities comply by size of municipality. Values incolumns 3-6 represent relative frequencies (i.e., the percentage of municipalities that comply with a percentage of items within those

n

intervals). 2. "H" refers to the Herfindahl-H index: H = ^ ( p ? ) = 1 .


24(11. 2-14


Table 5Chi Square Tests Results for Concentration

Subsystems x^ (*)Budgetary accounting 3.5794Financial accounting 16.3903

Note: (*) x^ = 7.81 with 5% significance level to 3dfs.

HI: The level of compliance is higher in largermunicipalities.

The literature also notes that profitability is associ-ated with the extent of compliance (Street & Gray, 2002),meaning that an adequate financial condition makes iteasier to comply with accounting standards. The relianceon debt negatively affects implementation and compli-ance (Christiaens, 1999). Taking into account the datalimitations, we include three variables to test the influ-ence of resource availability. The first is budgetarysurplus (revenues minus expenditures). The second is thenumber of financial transfers received from the centralgovernment (grants). The third is financial dependency,which is the proportion of grants to local governmentrevenues. In the first two cases, the expected sign is posi-tive, while in the last it is negative. The amount of publicdebt is excluded because municipalities are bound bydebt limitations, which means that the inclusion of thisvariable would not reveal any real influence of this factor.Many municipalities have reached their limit meaningthat the measurement of the variable would be truncated,severely biasing results.

H2: The financial condition of a municipality is relatedpositively to its level of compliance.

Environmental factors that the municipality has toface are also relevant in explaining differences in con-formity among municipalities. Arguably, a democrati-cally elected local government has to respond to theneeds of its citizens. In this context, studies on the deter-minants of what is called fiscal transparency (Alt, Lassen,& Rose, 2006) have pointed to political competition andspecifically to the nature or characteristics of the localelectorate. In the Portuguese political context, urbanareas are characterized by more educated electorates thatusually vote for more progressive and reformative politi-cal agendas. Accordingly, we hypothesize that this typeof electorate should be more sensitive to reforms thatimprove on fiscal transparency such as the implementa-tion of accounting reforms. We posit that more densely ,populated or urban areas are more prone to require higherlevels of compliance and therefore include a variable to

account for population density and a dummy variablesignifying the two largest metropolitan areas of Lisbonand Porto.

H3: Municipalities of the urban and more populatedareas have higher levels of compliance.

Christiaens (1999) describes the influence of themunicipal accounting staff on level of compliance. Weexamined the association between number of municipalemployees (per thousand inhabitants) and level of com-pliance of the municipality. Since the total number ofemployees is almost perfectly correlated with population,inclusion of total employees simply serves as an addi-tional proxy for size of municipality.

H4: Personnel size of a municipality is positivelyassociated with levels of municipality compliance.

Finally, we speculate that geographical proximitymay infiuence the adoption of the new accounting stan-dards. It is plausible to expect that the level of com-pliance is high in regions where the neighbouringmunicipalities also have high levels of compliance. Theprocess of propagation or emulation of reforms is notnew and is more likely in the cases of geographicalproximity. We include a variable that measures theaverage compliance level in the region of the municipal-ity. Since there are 18 administrative regions (called dis-tricts) that serve as the basis for the organization of thenational government administrative, we use the averagecompliance in the district. We expect it to positivelyaffect the level of compliance of a municipality.

H5: The average district level of compliance positivelyaffects the level of compliance of a municipality.

We test the above hypotheses using multiple linearregression analysis. The appendix shows the descriptivestatistics used in this analysis and the expected signsderived from each of our hypotheses. Results are shownin Table 6.

Although size does seem to have an effect, it iscontrary to our predictions. Given the other factors, theresults indicated that the larger the population, the lowerthe average level of compliance. These results differfrom the simple relations between size and complianceobserved in Table 2. Figure 1 shows that the three groupsconsidered to obtain partial levels of compliance havevery unbalanced numbers of municipalities. While thesmall sized group includes 91 municipalities and themedium sized group includes 62, the large sized groupincludes only 22 municipalities. This unbalanced group-ing of municipalities may produce artificial results. This


24rn. 2-14


Table 6Ordinary Least Squares Results for the Determinantsof Totai Compliance

VariablesLog population

Budgetary surplus

Grants

Financial dependency

Population density

Metropolitan areas

Municipal employees per thousandinhabitantsAverage compliance in the district

Constant

F(8, 160)Prob > FR-squaredRoot MSE

-4.408158***(1.348004)6.66e-07*

(3.53e-07)l.64e-06***

(6.08e-07)-3.391186(2.355998)-.0002731(.0004051)1.868679**(.9154943)-.0535108***(.0169436)1.02518***(.2155479)

18.93614***(6.936853)5.560.00000.16752.9387

Notes: 1. Total compliance is defined as INDEX = 2\p< where

p, = 1 for every parameter complied with and 0 otherwise (m£ 26). 2. Values in parentheses are standard errors; 3. *: p <.05; **: p < .01; ***: p < .001, two-tailed tests of statisticalsignificance; 4. To account for the potential problems of het-eroskedasticity, robust standard errors were computed using thecommand robust of the Stata statistical package.

potential problem is avoided with this regression analy-sis, in which the cases are treated individually. In addi-tion, one of the strengths of multivariate analysis is thatit allows control for other infiuences or explanations thatvery often modify a relation previously seen in bivariateanalyses. In more substantive terms and in contrast tosome studies, it seems that larger municipalities havedifficulties complying with accounting standards possi-bly due to organizational complexity, conservative prac-tices, and aversion to change.

The effect of financial conditions is as expected.Both the amount of grants received and the magnitudeof the budgetary surplus have positive effects on the levelof compliance. The level of financial dependency seemsto have no bearing.

The hypothesis regarding the effect of urban areas(H3) is partially confirmed. The municipalities that

belong to the metropolitan areas of Lisbon and Portohave significantly higher levels of compliance. Thisappears to be in line with the notion that municipal deci-sion-makers in these areas respond to electorates thatare presumably more demanding in terms of the qualityof information they receive, including financialinformation.

Finally, the results also suggest that a diffusion orpropagation of accounting practices may be present. Onaverage, the level of compliance is higher where thelevels of compliance of neighbouring municipalities arealso high. This finding is consistent with the diffusion ofinnovation.

Table 7 shows similar regression results forpartial compliance levels: budgetary, financial, andmanagement.

The results of partial compliance levels are substan-tively similar to those discussed above but weaker inexplanatory power, with the exception of financial com-pliance. This finding is driven by the fact that the depen-dent variables measuring budgetary and managementcompliance exhibit considerably less variation than thedependant variable measuring financial compliance. Asseen in Tables 1 and 2, the levels of budgetary compli-ance are very high, suggesting that the variation is insuf-ficient to obtain statistically significant coefficients. Incontrast, levels of management compliance are very low,meaning again that variation is low. The table of descrip-tive statistics in the appendix shows that in both casesthe standard deviation of these two variables is verylow.

Since there is much more variation in the levels offinancial compliance, the regression results are muchstronger than in the other two cases. The findings are verysimilar. They suggest the same type of effect in the caseof size, financial conditions, and the diffusion process.

Discussion

In this section, we present some reasons that mightjustify the observed levels of compliance with the Portu-guese local government's new accounting system. Incontrast to the previous section where we explained dif-ferences in compliance across municipalities, here weexplain general reasons that may account for the delayin the implementation of the accounting system. Sincethese reasons apply equally to all municipalities, they donot explain differences in implementation. They are asfollows:

- No legal establishment for municipalities' accounts tobe audited - Audits are carried out by the Court of

Copyright © 2007 ASAC. Published by John Wiley & Sons, Ltd. 10Can J Adm Sci

24(11. 2-14


Table 7Ordinary Least Squares Results for Determinants of Partial Compliance

Variables

Log population

Budgetary surplus

Grants

Financial dependency

Population density

Metropolitan areas

Municipal employees/1000

Compliance in the district

Constant

F(8, 160)Prob > FR-squaredRoot MSE

Budgetary compliance

-.3720859(.4679755)

-1.76e-08(1.24e-07)6.17e-08

(2.05e-07)-.7186432(.7745403).0000728

(.0000949)-.0927911(.3393132)

-.0064971(.0077705).2870186***

(.0795698)3.733096

(2.540107)4.590.00000.07891.0625

Financial compliance

-3.155892***(.8790023)5.83e-O7**

(2.24e-07)1.3le-06***

(4.06e-07)-2.303959(1.548764)-.0002854(.0002744)1.403941**(.5916822)

-.0374101***(.0109209).5542595***

(.1452448)12.90161***(4.593188)5.060.00000.16031.8983

Management compliance

-.8801804***(.3265608)1 .OOe-07

(7.98e-08)2.76e-07*

(1.41e-07)-.3685837(.5474944)

-.0000605(.0000776).5575294***

(.1921043)-.0096035(.0043199)**.1839019***

(.0566169)2.301442

(1.785142)3.630.00070.1343

.67991

Notes: 1. Total compliance is defined as INDEX = ^ p i where p , = 1 for every parameter complied with and 0 otherwise (m £ 26).

2. Values in parentheses are standard errors; 2) *: p < .05; **: p < .01; ***: p < .001, two-tailed tests of statistical significance; 3. Toaccount for the potential problems of heteroskedasticity, robust standard errors were computed using the command robust of the Statastatistical package.

Accounts and the Finance General Inspection only forthe purpose of scrutinizing legal form and fiscal regu-larity. Audits do not assess compliance with the newaccounting system requirements in terms of policiesand practices.

- No penalties (neither fiscal nor economic) for thosethat do not comply with the new system requirements.- Although compliance is required by the Court ofAccounts in terms of types of financial and budgetarystatements to be presented for inspection, noncompli-ance with accounting policies and reporting contentsis acceptable.

- Lack of further clarification, namely a conceptualframework that supports the implementation ofaccrual-based accounting - As in other countries,accrual accounting and financial reporting has beenpresented as a "good thing" bringing large advantagesto the municipalities' accounts. The business account-ing policies and reporting model have not been adapted

to municipalities' specificities, requiring all assets tobe listed in the balance sheet. Entities have facedserious problems in inventorying and valuing allassets, as well as applying the accrual concept. On theother hand, the main users, namely local politicians,have problems in understanding the usefulnessof such accrual-based financial statements fordecision-making.Problems with qualification levels of the accountingstaff and professional involvement of the chief finan-cial officer - Lack of training in accrual accountingtechniques and lack of involvement in professionalaccounting organizations (i.e., bookkeepers who arenot familiar with business accounting) might have hin-dered the adoption of new practices such as double-entry accounting and the accrual principle. Furthermore,contrary to what happens in the business sector, thereis not yet a career of certified accountant for the publicadministration context.


24fn. 2-14


- Cost/management accounting difficulties - There is alack of software availability, particularly with func-tional classification of costs. Also, because there is afull costing requirement, there are problems with theallocation of certain types of costs. Additionally, thereis no requirement to use standard costs against whichactuals could be compared in order to assesseffectiveness.

- Hesitancy to hire a consulting firm - Given the auton-omy in decision-making of these firms, as well as themunicipalities' financial capacity, few municipal-ities can easily hire external consulting firmsthat would help implement the new accountingsystem.

Conclusion

We have assessed compliance of public officesthroughout Portugal with respect to the standards intro-duced by the OAP-LG. Through reviewing and analyz-ing the information reflected in the accounts andmanagement reports of a sample of Portuguese munici-palities for 2003, we have shown that the implementationof the new accounting system is still far from complete.The average level of total compliance with the newsystem is 58%. We considered the three accounting sub-systems separately and found that their implementationevolved at different speeds. The highest average partialcompliance level is for the budgetary accounting subsys-tem (87%), followed by the financial accounting systemwith a level of 54%, and then by the managementaccounting system with a level of only 26.7%.

We have also assessed the cross-sectional diversityof practices and information within both budgetary andfinancial accounting subsystems. The diversity of prac-tices and information is lower for those concerning bud-getary accounting, regardless of municipality size.

We have also offered explanations for differencesacross municipalities in the levels of compliance. Ourregression analyses show some interesting results. Aboveall, size of municipality has a negative effect on compli-ance. Larger municipalities appear to have more diffi-culty in complying with accounting standards, whichmay be due to organizational complexity, conservativepractices, and aversion to change. The differences incompliance are also explained by the financial conditionsof the municipality, particularly the availability of finan-cial resources. The number of grants received, as well ashaving a budgetary surplus, has positive effects on com-pliance. Also of interest is the fact that municipalities inthe Porto and Lisbon metropolitan areas have signifi-cantly higher levels of compliance. High levels of com-

pliance are found where levels of compliance inneighbouring municipalities are also high.

The substantive relevance of these results is twofold.Firstly, there is still a significant gap between intentionand practice regarding the reform of Portuguese localgovernment accounting. Much is yet to be done, particu-larly at the level of management accounting. This is anaccounting subsystem that is absolutely necessary toproduce detailed information on the costs of public activ-ity and ultimately, to evaluate the efficiency of the use ofpublic resources. In more general terms, citizens andother users need more quality in local governmentaccounting information so that this information isreliable and comparable in a way that allows forevaluation.

Secondly, a number of the explanations for the delayin compliance are municipality-specific: some munici-palities simply take more time than others in the imple-mentation of reforms. It is, however, important, thatmunicipalities adapt quickly due to the increasing paceof additional reporting requirements. A new challenge tomunicipalities is forthcoming in the form of a LocalFinance Law introducing new requirements with respectto borrowing, intergovernmental grants, and consolidatedaccounts. Municipal governments need to be concernedwith fully implementing the new accounting system.

Our research also opens some avenues for futureresearch. Because this is the first attempt to study thePortuguese local government case, the present analysiscould be refined. To compute the compliance index, weassumed that the 26 accounting parameters are equallyimportant and therefore have the same weight. This isobviously a simplification justified by the exploratorynature of this work. Future work should carefully con-sider different weights.

Notes

1 This is a law-based general accounting plan similar toothers existing in France, Belgium and Spain. In Portugalit is the fundamental set of instructions for governmentalaccounting practices and it contains: a standard decimalizedchart of accounts, instructions relating to the presentationof uniform published financial and budgetary statements,and standard definitions of items and their valuationmethods.

2 This was a military coup that restored democracy after aforty-year dictatorship.

3 According to the Constitution (article 214, n.l). The Courtof Accounts is the 'supreme body of supervision of thelegality of public expenses and of judgement of the report-ing statements that the law decides to submit to it'. It super-vises the legality and financial regularity of the publicexpenses and revenues. Therefore, the State and its ser-


24(11. 2-14


vices, autonomous regional governments, local govern-ment, public institutes and social security institutions, aresubjected to the jurisdiction and financial control of TheCourt of Accounts.

The number of entities under this condition surpasses11,000 and the reporting statements to be submitted toauditing/inspection are defined in two resolutions: for LocalGovernment, Resolution 04/2001 and for all the other enti-ties. Resolution 01/2004.

4 Similar methodologies were followed by Christiaens (1999)and Torres and Pina (2003), among others. The latter addi-tionally discusses other approaches (within the businesscontext) that weigh each item in the index differently, high-lighting problems of subjectivity associated with establish-ing the importance of each item while considering differentusers and objectives of the information provided.

5 Following the suggestion of an anonymous reviewer, wesegmented the parameters (Table 1) into those that representlong-term asset accounting (financial accounting parame-ters 1^) as opposed to the other accrual accounting param-eters (financial accounting parameters 5-10) and wecomputed an index for each set separately. The expectationwas to find compliance for financial accounting parameters5-10 to be significantly higher than those for 1—4. Never-theless, the results showed a different picture. The compli-ance for long-term assets parameters (1-4) is significantlyhigher than those for accrual accounting parameters(5-10).

References

Allen, A., & Sanders, G. (1994). Financial disclosures in USmunicipalities: has the governmental accounting standardsboard made a difference.'' Financial Accountability andManagement, IO{3), 175-193.

Alt, J., Lassen, D., & Rose, S. (2006). The Causes of FiscalTransparency: Evidence from the American States. Manu-script in preparation.

Carvalho, J.C, & Jorge, S. (2003, April). Governmentalaccounting in Portugal and the information disclosed oncash and accruals bases: the case of Porto municipality.Paper presented at the 26"' Annual Congress of the Euro-pean Accounting Association, Seville.

Christiaens, J. (1999). Financial accounting reforms in Flemishmunicipalities: an empirical investigation. FinancialAccountability and Management, 75(1), 21-40.

Femandes, M.J., & Carvalho, J.B.C. (2004, July). The RecentDevelopment in Public Accounting in Portugal: resultsafter seven years of reform implementation.. Paper pre-sented at the 2"'' International Conference in Accountingand Finance in Transition, Kavala, Greece.

Guthrie, J. (1998). Application of accrual accounting in theAustralian Public Sector - Rhetoric or Reality? FinancialAccountability and Management, I4{\), 1-19.

Lapsley, 1., Brown, T., Jackson, A., Oldfield, R., & Pong, C.(2003). The transformation of the Public Sector: the roleof accounting in sustaining change. The Institute of Char-tered Accountants of Scotland.

Jorge, S.M. (2003). Local Government accounting in Portugalin Comparative-International Perspective., PhD Thesis.Birmingham Business School. The University ofBirmingham.

Jorge, S.M. (2004, July) Local Government accounting in Por-tugal: recent changes and future tendencies. Paper pre-sented at the 2nd International Conference on Accountingand Finance in Transition, Kavala, Greece.

Pina Martinez, V., & Torres, L. (1995, May). ComparativeStudy of the Governmental Financial Reports in six Coun-tries. Paper presented at the 5th CIGAR Conference,Paris.

Street, D., & Gray, S. (2002). Factors Influencing the Extent ofCorporate Compliance with International AccountingStandards: Summary of a Research Monograph. Journalof International Accounting Auditing and Taxation, / /(1),51-76.

Torres, L., & Pina Martfnez, V. (2003, April) 'Local Govern-ment financial reporting in the USA and Spain: a Com-parative Study. Span/.s/jyoMmo/o/'H/ja/icefl/irfAccoM/;///)^,115, 153-183.

Legislation:

Constituifao da Reptiblica Portuguesa (2000) - actualizada deacordo com a Lei Constitucional n°l/97, de 20 de Setem-bro; Almedina; Coimbra; Outubro.

Instruction 671/2000 - Assessment and Inventory of PublicProperty.

Instruction 794/2000 - Official Accounting Plan forEducation.

Instruction 898/2000 - Official Accounting Plan for Health.Law n°42/98 - Local Finance Law.Law-decree 68/98 - Public Administration Accounting Stan-

dardization Commission.Law Decree n°54-A/99 - Official Accounting Plan for Local

Government [OAP-LG].Law Decree n°l2/2002 - Official Accounting Plan for Social

Security.Law Decree 26/2002 - Economic budgetary classification for

revenues and expenditures.Resolution 4/2001 - Rules to prepare the accounts statetrtents

for The Court of Accountants.


24fn. 2-14

CONFORMITY AND DIVERSITY OF ACCOUNTING AND FINANCIAL REPORTING PRACTICES DA COSTA CARVALHO ETAL

Appendix

Descriptive Statistics and Expected signs

Variable

Total complianceBudgetary complianceFinancial complianceManagementcomplianceLog populationBudgetary surplusGrantsFinancial dependencyPopulation densityMetropolitan areasMunicipalemployees/1000Compliance in the district

Obs.

170170170170

170170170170170170170

170

Mean

15.052946.1117657.6058821.335294

4.326175-09273.91227811.4374843417.4022.164705916.69439

15.03412

Std. Dev

3.142121.0793282.018544.7128916

.500038510024701077360.18086881084.703.372010710.96734

.9653835

Min.

3300

3.431042-.28e+O7382015.10070976.5602.081332

13.1

Max

207

112

5.745697555238.1l.O8e+O7.90299497597.17199.0502

17.7

Exp. Sign

n.a.n.a.n.a.n.a.

+++-+++

+


24(11.2-14

Documents

Potu Gal